Data integration for measuring emigration flows: a new ...• Measuring migration movements of populations in different countries is a challenging task. • Administrative sources

Data integration for measuring emigration flows: a new approach in the Italian case Enrico Tucci [email protected]; Gianni Corsetti [email protected]; Francesca Licari [email protected]; Valeria Tomeo [email protected] OECD-IOM-UNDESA International Forum on Migration Statistics 15-16 January 2018, Paris

Aim of the work

Establish a coherent and consistent data base that contains sufficiently detailed, up-to-date and accurate information by mean of data integration from several sources (Frans Willekens, European Journal of Population 10: 1-42, 1994) The outputs of the process should: 1. meet the needs of users (relevance) 2. accurately and reliably portray reality (accuracy and reliability) 3. be consistent internally, over time and comparable between regions and

countries; it is possible to combine and make joint use of related data from different sources (coherence and comparability)

Migration statistics: two issues

1. Conceptual issue: difference between administrative concept and statistical definition (gap between supply and demand) . Illegal staying in the country 12 months rule

2. Measurement issue: how to measure emigration flows (reliability of the official figures) “missing” flows (excluded from the official statistics) deregistration ex officio (excluded from the official statistics)

1 Measurement issue: missing flows. Record linkage between Micro Demographic Account and Population Registers

Pop Census 09/10/2011 Resulting Pop 01/01/2016

Resu

lting Po

p 0

1/0

1/2

01

6

Pop

Register 0

1/0

1/2

01

6

Data integration within the same source

Do they exit the population by emigration or death?

Administrative sources and presence on the territory

Availability of many administrative sources: labor and education registers, consular register, tax returns register, earnings, retired, and non-pension benefits registers, permits to stay.

1 Measurement issue: missing flows. Data integration with other administrative sources

At the end of the process more than 250 thousand individuals have been identified as emigrant and will be included in the official statistics

Missing flows by event of exit Administrative sources A.V.

a) Deaths tax returns register 22,475

b) National emigrants consular register 49,280

c) Other emigrants no signals 135,984

d) Potential emigrants with signals of labour/study 88,658

Total 296,397

2 Measurement issue: deregistration ex officio

From an administrative to a demographic balance

From the Census date (09/10/2011) to 01/01/2016 there have been 449,704 deregistrations ex officio.

293,988 individuals are not present in the stock population data of the following years

They have likely left the country without informing the administrative accountable office of their departure.

Those deregistrations are included in the demographic balance as «deregistration for other reasons» and not considered as emigration flows

Information to be estimated

All the main information about flow (‘who’, ‘when’, ‘where’) are known in the case of people that declare their departure

‘Who’ is known. ‘When’ and ‘where’ are unknown in the case of deregistration ex officio

Emigration to abroad

From the Population Census (09/10/2011) to 01/01/2016 there have been 409,758 emigrations to abroad. Every individual emigration flow has been associated with the respective Administrative Signal on the territory. The underlying idea is that there is a relationship between the Date of emigration (Date EM) and the date of the Latest Administrative Signal (Date LAS) and that the two Dates are related

[Diff EMLAS = Date EM – Date LAS] (in months)

Deregistration ex officio

Date of deregistration ex officio (Date DER) of a person does not correspond to the date of occurrence of the emigration (Date EMIG). Also the individual deregistration ex officio has been associated with the Administrative Signal on the territory.

Diff DRLAS= Date DER – Date LAS

Deregistration ex officio

Diff EMLAS distribution (by CTZ and CTB) has been used to estimate the date of occurrence of the emigration (Date EMIG*) for those deregistration carried out ex officio Diff EMLAS= Date EMIG – Date LAS (normally distribuited) Diff DRLAS= Date DER – Date LAS Random (Diff EMLAS)= Date EMIG* – Date LAS

Date EMIG*= Random(Diff EMLAS) + Date LAS

Results (1)

Results (2)

Deregistration ex officio and other missing emigration flows by year of

estimated emigration and Citizenship

Years (estimated) Nationals Not Nationals A.V.

2011 (from 09/10) 11,930 36,514 48,444

2012 34,202 87,204 121,406

2013 36,142 90,470 126,612

2014 42,082 91,571 133,653

2015 45,041 92,752 137,795

Total 169,397 398,513 567,910

Results (3)

Immigration, emigration flows and Net migration (observed and estimated*)

by Citizenship - Year 2016

Immigration Emigration Net MIG Emigration* Net MIG*

Nationals 30,052 102,259 -72,207 147,300 -117,248

Non Nationals 250,026 44,696 205,33 137,450 112,576

Total 280,078 146,955 133,123 284,750 -4,672

EstimatedObservedCitizenship

First Conclusions

• Measuring migration movements of populations in different countries is a challenging task.

• Administrative sources are an extremely powerful tool and a potential source of information for emigration statistics.

• The Italian case provided empirical evidence of the issues to be confronted and the challenges to the use of a data integration for improving the quantity and quality of data on emigration.

• A coherent and consistent data base that contains detailed, up-to-date and accurate information allows to study migration through a longitudinal approach (Return Migration, Circular Migration)